XPK Start: Mon Apr 20 19:23:04 UTC 2026 2026-04-20 19:23:08.415279: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:467] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered WARNING: All log messages before absl::InitializeLog() is called are written to STDERR E0000 00:00:1776712988.427722 10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered E0000 00:00:1776712988.431470 10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered W0000 00:00:1776712988.442952 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776712988.442972 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776712988.442974 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776712988.442976 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. 2026-04-20 19:23:27.739818: E external/local_xla/xla/stream_executor/cuda/cuda_platform.cc:51] failed call to cuInit: INTERNAL: CUDA error: Failed call to cuInit: UNKNOWN ERROR (303) I0420 19:23:28.275353 135546001930048 max_utils.py:273] Attempting to initialize the jax distributed system... INFO:2026-04-20 19:23:37,318:jax._src.distributed:140: Starting JAX distributed service on [::]:8482 I0420 19:23:37.318320 135546001930048 distributed.py:140] Starting JAX distributed service on [::]:8482 INFO:2026-04-20 19:23:37,320:jax._src.distributed:157: Connecting to JAX distributed service on mt-04-int8-g3bt9-slice-job-0-0.mt-04-int8-g3bt9:8482 I0420 19:23:37.320671 135546001930048 distributed.py:157] Connecting to JAX distributed service on mt-04-int8-g3bt9-slice-job-0-0.mt-04-int8-g3bt9:8482 I0420 19:23:38.368257 135546001930048 max_utils.py:284] Jax distributed system initialized! I0420 19:23:45.301772 135546001930048 max_utils.py:800] System Information: Jax Version: 0.8.1 I0420 19:23:45.301879 135546001930048 max_utils.py:801] System Information: Jaxlib Version: 0.8.1 I0420 19:23:45.301920 135546001930048 max_utils.py:802] System Information: Jax Backend: PJRT C API TFRT TPU v6 lite Built on Nov 12 2025 14:16:36 (1762985796) cl/831091709 I0420 19:23:45.301955 135546001930048 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0420 19:23:45.911969 135546001930048 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0420 19:23:45.912597 135546001930048 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0420 19:23:45.912905 135546001930048 checkpointing.py:688] Setting up checkpoint logger... I0420 19:23:45.912960 135546001930048 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True I0420 19:23:45.913004 135546001930048 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None I0420 19:23:45.913350 135546001930048 base_pytree_checkpoint_handler.py:415] Created BasePyTreeCheckpointHandler: use_ocdbt=True, use_zarr3=True, pytree_metadata_options=PyTreeMetadataOptions(support_rich_types=False), array_metadata_store=<orbax.checkpoint._src.metadata.array_metadata_store.Store object at 0x7b46bf1919d0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0420 19:23:49.019981 135546001930048 checkpointing.py:266] Enabling policy for fixed interval checkpointing. I0420 19:23:49.020341 135546001930048 checkpoint_manager.py:709] [process=0][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b32883edc70>}, handler_registry=None I0420 19:23:49.020578 135546001930048 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b32883edc70>` for item "items" and save args `<class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeSaveArgs'>` and restore args `<class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>` to `_handler_registry`. I0420 19:23:49.020634 135546001930048 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b32883f30e0>` for item "metrics" and save args `<class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>` and restore args `<class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>` to `_handler_registry`. I0420 19:23:49.020671 135546001930048 composite_checkpoint_handler.py:505] Initialized registry DefaultCheckpointHandlerRegistry({('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeSaveArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b32883edc70>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b32883edc70>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b32883f30e0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b32883f30e0>}). I0420 19:23:49.021201 135546001930048 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33 I0420 19:23:49.021280 135546001930048 async_checkpointer.py:177] [process=0][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7b32881b0180> timeout: 600 secs and primary_host=0 for async checkpoint writes I0420 19:23:50.163970 135546001930048 checkpoint_manager.py:1818] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints I0420 19:23:50.166248 135546001930048 checkpoint_manager.py:929] [process=0][thread=MainThread] CheckpointManager created, primary_host=0, CheckpointManagerOptions=CheckpointManagerOptions(save_interval_steps=1, max_to_keep=None, keep_time_interval=None, keep_period=None, should_keep_fn=None, best_fn=None, best_mode='max', keep_checkpoints_without_metrics=True, step_prefix=None, step_format_fixed_length=None, step_name_format=None, create=True, cleanup_tmp_directories=False, save_on_steps=frozenset(), single_host_load_and_broadcast=False, todelete_subdir=None, todelete_full_path=None, enable_background_delete=False, read_only=False, enable_async_checkpointing=True, async_options=None, multiprocessing_options=MultiprocessingOptions(primary_host=0, active_processes=None, barrier_sync_key_prefix=None), should_save_fn=None, file_options=FileOptions(path_permission_mode=None), save_root_metadata=True, temporary_path_class=None, save_decision_policy=FixedIntervalPolicy(interval=10), preservation_policy=LatestN(n=None), prevent_write_metrics=False, enable_should_save_is_saving_in_progress_check=True, enable_per_process_directory_creation=False, lightweight_initialize=False), root_directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7b32883ecc20> I0420 19:23:50.166362 135546001930048 checkpointing.py:302] Checkpoint manager created! I0420 19:23:51.357161 135546001930048 nnx_wrappers.py:455] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0420 19:23:51.357267 135546001930048 nnx_wrappers.py:455] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0420 19:23:51.736978 135546001930048 attentions.py:1084] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0420 19:23:51.737073 135546001930048 attentions.py:1084] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0420 19:23:51.753036 135546001930048 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0420 19:23:51.753113 135546001930048 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0420 19:23:51.833396 135546001930048 attentions.py:1150] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0420 19:23:51.833489 135546001930048 attentions.py:1150] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0420 19:23:51.849579 135546001930048 attentions.py:1151] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0420 19:23:51.849641 135546001930048 attentions.py:1151] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0420 19:23:51.865647 135546001930048 attentions.py:1152] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0420 19:23:51.865710 135546001930048 attentions.py:1152] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0420 19:23:51.890007 135546001930048 attentions.py:1193] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0420 19:23:51.890087 135546001930048 attentions.py:1193] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0420 19:23:51.974930 135546001930048 linears.py:541] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0420 19:23:51.975025 135546001930048 linears.py:541] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0420 19:23:52.595909 135546001930048 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint I0420 19:23:52.596024 135546001930048 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint. fsdp: 32 I0420 19:23:55.296477 135546001930048 maxtext_utils.py:1740] params/params/decoder/decoder_norm/scale Shape: float32[2048] Logical: PartitionSpec('norm',) Physical: (None,) I0420 19:23:55.296618 135546001930048 maxtext_utils.py:1740] params/params/decoder/layers/mlp/wi_0/kernel Shape: float32[2048,16,7168] Logical: PartitionSpec('embed', 'layers', 'mlp') Physical: ('fsdp', None, None) I0420 19:23:55.296675 135546001930048 maxtext_utils.py:1740] params/params/decoder/layers/mlp/wi_1/kernel Shape: float32[2048,16,7168] Logical: PartitionSpec('embed', 'layers', 'mlp') Physical: ('fsdp', None, None) I0420 19:23:55.296735 135546001930048 maxtext_utils.py:1740] params/params/decoder/layers/mlp/wo/kernel Shape: float32[7168,16,2048] Logical: PartitionSpec('mlp', 'layers', 'embed') Physical: (None, None, 'fsdp') I0420 19:23:55.296791 135546001930048 maxtext_utils.py:1740] params/params/decoder/layers/post_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: PartitionSpec('norm', 'layers') Physical: (None, None) I0420 19:23:55.296833 135546001930048 maxtext_utils.py:1740] params/params/decoder/layers/pre_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: PartitionSpec('norm', 'layers') Physical: (None, None) I0420 19:23:55.296892 135546001930048 maxtext_utils.py:1740] params/params/decoder/layers/self_attention/key/kernel Shape: float32[2048,16,16,128] Logical: PartitionSpec('embed', 'layers', 'kv_heads', 'kv_head_dim') Physical: ('fsdp', None, None, None) I0420 19:23:55.296948 135546001930048 maxtext_utils.py:1740] params/params/decoder/layers/self_attention/out/kernel Shape: float32[16,16,128,2048] Logical: PartitionSpec('heads', 'layers', 'kv', 'embed') Physical: (None, None, None, 'fsdp') I0420 19:23:55.296992 135546001930048 maxtext_utils.py:1740] params/params/decoder/layers/self_attention/query/kernel Shape: float32[2048,16,16,128] Logical: PartitionSpec('embed', 'layers', 'q_heads', 'kv') Physical: ('fsdp', None, None, None) I0420 19:23:55.297033 135546001930048 maxtext_utils.py:1740] params/params/decoder/layers/self_attention/value/kernel Shape: float32[2048,16,16,128] Logical: PartitionSpec('embed', 'layers', 'kv_heads', 'kv_head_dim') Physical: ('fsdp', None, None, None) I0420 19:23:55.297101 135546001930048 maxtext_utils.py:1740] params/params/decoder/logits_dense/kernel Shape: float32[2048,32000] Logical: PartitionSpec('embed', 'vocab') Physical: ('fsdp', None) I0420 19:23:55.297167 135546001930048 maxtext_utils.py:1740] params/params/token_embedder/embedding Shape: float32[32000,2048] Logical: PartitionSpec('vocab', 'embed') Physical: (None, 'fsdp') I0420 19:23:57.046753 135546001930048 train.py:158] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0420 19:23:57.046850 135546001930048 train.py:158] train/xent Physical: float32[32,2048]............................................ ('fsdp', None). I0420 19:23:57.061794 135546001930048 train.py:165] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0420 19:23:57.061851 135546001930048 train.py:165] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None). I0420 19:24:10.917026 135546001930048 max_utils.py:791] Total memory size: 1.8 GB, Output size: 0.4 GB, Temp size: 1.4 GB, Argument size: 0.4 GB, Host temp size: 0.0 GB. I0420 19:24:10.923379 135546001930048 metric_logger.py:289] number parameters: 1.104 billion I0420 19:24:26.940361 135546001930048 checkpointing.py:794] Waiting for step 0 to finish before checkpoint... I0420 19:24:27.180029 135546001930048 checkpointing.py:798] Waited 0.23964571952819824 seconds for step 0 to finish before starting checkpointing. I0420 19:24:27.182541 135546001930048 checkpoint_manager.py:2013] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0420 19:24:27.184391 135546001930048 checkpoint_manager.py:1518] [process=0] Saving checkpoint at step 0 I0420 19:24:27.187293 135546001930048 async_checkpointer.py:452] [process=0] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/0. I0420 19:24:27.507863 135546001930048 signaling_client.py:364] Using JaxDistributedSignalingClient I0420 19:24:27.775562 135413456951040 atomicity.py:137] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/0 I0420 19:24:27.814458 135546001930048 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array. I0420 19:24:27.814552 135546001930048 replica_slices.py:424] Transferring arrays to host memory with options: use_replica_parallel=True, min_slice_bytes_for_replica_parallel=None, max_replicas_for_replica_parallel=None, enable_pinned_host_transfer=False I0420 19:24:28.068590 135546001930048 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.255226s I0420 19:24:28.069483 135546001930048 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 2.785 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.5538065433502197 s) (per-host) I0420 19:24:28.069556 135546001930048 base_pytree_checkpoint_handler.py:737] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.553898s (batch_requests_ready=0.293783s, total_serialization_initiated=0.259307s, others=0.000808s) I0420 19:24:28.069637 135546001930048 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.558095s (all_items=0.000018s, per_item={'items': '0.00001836'}, temp_paths=0.558077) I0420 19:24:28.070579 135413965965056 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started. I0420 19:24:28.070736 135546001930048 async_checkpointer.py:561] Finished blocking save. Time taken: 0.886272s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/0. I0420 19:24:28.074443 135546001930048 checkpoint_manager.py:1566] [process=0][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0420 19:24:28.074735 135413448558336 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0420 19:24:28.074927 135546001930048 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776713067.1825228, 'wait_for_prev_duration_secs': 7.295608520507812e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776713067.184431, 'checkpointer_blocking_duration_secs': 0.8864161968231201, 'get_old_steps_start_time': 1776713068.0708663, 'get_old_steps_duration_secs': 2.5987625122070312e-05, 'checkpoint_manager_blocking_start_time': 1776713067.1807094, 'checkpoint_manager_blocking_duration_secs': 0.8941769599914551} I0420 19:24:28.075044 135546001930048 checkpointing.py:409] Started an asynchronous checkpoint save for step 0 I0420 19:24:28.075121 135546001930048 max_utils.py:750] Memstats: After params initialized: I0420 19:24:28.075175 135546001930048 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_0(process=0,(0,0,0,0)) I0420 19:24:28.075212 135546001930048 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_1(process=0,(1,0,0,0)) I0420 19:24:28.075242 135546001930048 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_4(process=0,(0,1,0,0)) I0420 19:24:28.075268 135546001930048 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_5(process=0,(1,1,0,0)) I0420 19:24:28.372864 135546001930048 metric_logger.py:185] completed step: 0, seconds: 15.686, TFLOP/s/device: 0.866, Tokens/s/device: 130.560, total_weights: 65536, loss: 10.874 I0420 19:24:28.375666 135546001930048 metric_logger.py:269] To see full metrics 'tensorboard --logdir=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/tensorboard/' I0420 19:24:28.791927 135546001930048 metric_logger.py:185] completed step: 1, seconds: 1.431, TFLOP/s/device: 9.496, Tokens/s/device: 1431.328, total_weights: 65536, loss: 10.874 I0420 19:24:29.019488 135546001930048 metric_logger.py:185] completed step: 2, seconds: 0.420, TFLOP/s/device: 32.361, Tokens/s/device: 4877.840, total_weights: 65536, loss: 10.262 I0420 19:24:29.248979 135546001930048 metric_logger.py:185] completed step: 3, seconds: 0.008, TFLOP/s/device: 1626.422, Tokens/s/device: 245152.023, total_weights: 65536, loss: 9.731 I0420 19:24:29.664556 135417189869312 atomicity.py:137] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/0/items I0420 19:24:52.141647 135413957572352 checkpoint.py:188] Wrote Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776713069398425361, 'commit_timestamp_nsecs': None, 'custom_metadata': {}}, json={"item_handlers": null, "metrics": {}, "performance_metrics": {}, "init_timestamp_nsecs": 1776713069398425361, "commit_timestamp_nsecs": null, "custom_metadata": {}} to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/0/_CHECKPOINT_METADATA I0420 19:24:52.143005 135546001930048 metric_logger.py:185] completed step: 4, seconds: 0.229, TFLOP/s/device: 59.407, Tokens/s/device: 8954.493, total_weights: 65536, loss: 9.272 I0420 19:24:52.152180 135546001930048 metric_logger.py:185] completed step: 5, seconds: 0.229, TFLOP/s/device: 59.234, Tokens/s/device: 8928.338, total_weights: 65536, loss: 8.887 I0420 19:24:52.379693 135546001930048 metric_logger.py:185] completed step: 6, seconds: 22.894, TFLOP/s/device: 0.593, Tokens/s/device: 89.456, total_weights: 65536, loss: 8.588 I0420 19:24:52.609213 135546001930048 metric_logger.py:185] completed step: 7, seconds: 0.007, TFLOP/s/device: 2013.803, Tokens/s/device: 303542.315, total_weights: 65536, loss: 8.380 I0420 19:24:52.838822 135546001930048 metric_logger.py:185] completed step: 8, seconds: 0.228, TFLOP/s/device: 59.474, Tokens/s/device: 8964.606, total_weights: 65536, loss: 8.251 I0420 19:24:53.067757 135546001930048 checkpointing.py:794] Waiting for step 9 to finish before checkpoint... I0420 19:24:53.068448 135546001930048 checkpointing.py:798] Waited 0.0007128715515136719 seconds for step 9 to finish before starting checkpointing. I0420 19:24:53.070580 135546001930048 checkpoint_manager.py:2024] [process=0][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0420 19:24:53.157241 2737 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0420 19:24:54.801545 135415572281088 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/0/items/array_metadatas/process_0 I0420 19:25:27.224565 135415563888384 base_pytree_checkpoint_handler.py:1232] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 34.071061s (commit=32.512616s, array_metadata_write=1.558445s) I0420 19:25:27.225611 135413965965056 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/gbytes_per_sec: 26.455 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 59.709903717041016 s) (per-host) I0420 19:25:27.225715 135413965965056 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 59.155027s. I0420 19:25:27.922224 135413965965056 checkpoint.py:228] Read Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776713069398425361, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} from gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/0/_CHECKPOINT_METADATA I0420 19:25:28.674293 135413957572352 checkpoint.py:247] Updated Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776713069398425361, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/0/_CHECKPOINT_METADATA I0420 19:25:28.675884 135413965965056 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000224s. I0420 19:25:29.289904 135413965965056 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389). I0420 19:25:30.226689 135413965965056 base_pytree_checkpoint_handler.py:1356] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.175512s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/0/items I0420 19:25:30.229839 135413965965056 atomicity.py:608] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/0/items I0420 19:25:30.650225 135413965965056 atomicity.py:608] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/0 I0420 19:25:31.956386 135413965965056 atomicity.py:790] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/0`. I0420 19:25:31.957240 135413965965056 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 64.772782s. directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/0 I0420 19:25:31.959961 135413965965056 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 63.889257s. I0420 19:25:31.960218 135413448558336 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0420 19:25:31.960294 135413448558336 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0420 19:25:31.960366 135413448558336 checkpoint_manager.py:2133] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0420 19:25:31.964144 135413448558336 checkpoint_manager.py:2142] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0420 19:25:31.964378 135546001930048 checkpoint_manager.py:2036] [process=0][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0. W0420 19:25:31.964518 135546001930048 checkpoint_manager.py:1458] Waiting for previous save to complete took 38.893943 seconds. If this number is high, consider checkpointing less frequently. I0420 19:25:31.966489 135546001930048 checkpoint_manager.py:1518] [process=0] Saving checkpoint at step 9 I0420 19:25:31.970124 135546001930048 async_checkpointer.py:452] [process=0] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/9. I0420 19:25:32.917614 135413965965056 atomicity.py:137] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/9 I0420 19:25:32.918538 135546001930048 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array. I0420 19:25:32.918639 135546001930048 replica_slices.py:424] Transferring arrays to host memory with options: use_replica_parallel=True, min_slice_bytes_for_replica_parallel=None, max_replicas_for_replica_parallel=None, enable_pinned_host_transfer=False I0420 19:25:32.945436 135546001930048 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.028162s I0420 19:25:32.946343 135546001930048 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 5.543 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2783079147338867 s) (per-host) I0420 19:25:32.946420 135546001930048 base_pytree_checkpoint_handler.py:737] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.278403s (batch_requests_ready=0.247043s, total_serialization_initiated=0.030530s, others=0.000830s) I0420 19:25:32.946504 135546001930048 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.282691s (all_items=0.000016s, per_item={'items': '0.00001597'}, temp_paths=0.282675) I0420 19:25:32.947422 135416661391104 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started. I0420 19:25:32.947574 135546001930048 async_checkpointer.py:561] Finished blocking save. Time taken: 0.981017s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/9. I0420 19:25:32.975426 135546001930048 checkpoint_manager.py:1566] [process=0][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0420 19:25:32.975701 135413448558336 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0420 19:25:32.975873 135546001930048 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776713093.0705469, 'wait_for_prev_duration_secs': 38.893943071365356, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776713131.9665284, 'checkpointer_blocking_duration_secs': 0.9811475276947021, 'get_old_steps_start_time': 1776713132.9476953, 'get_old_steps_duration_secs': 2.6464462280273438e-05, 'checkpoint_manager_blocking_start_time': 1776713093.0686805, 'checkpoint_manager_blocking_duration_secs': 39.90715765953064} I0420 19:25:32.975982 135546001930048 checkpointing.py:409] Started an asynchronous checkpoint save for step 9 I0420 19:25:32.976026 135546001930048 checkpoint_manager.py:2024] [process=0][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0420 19:25:34.711970 135412936865536 atomicity.py:137] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/9/items I0420 19:25:37.898133 135416678176512 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/9/items/array_metadatas/process_0 I0420 19:26:18.273325 135416669783808 base_pytree_checkpoint_handler.py:1232] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 42.034424s (commit=40.882954s, array_metadata_write=1.151470s) I0420 19:26:18.274338 135416661391104 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/gbytes_per_sec: 34.636 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 45.60627222061157 s) (per-host) I0420 19:26:18.274457 135416661391104 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 45.326924s. I0420 19:26:19.787224 135416661391104 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000322s. I0420 19:26:20.368676 135416661391104 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389). I0420 19:26:21.285114 135416661391104 base_pytree_checkpoint_handler.py:1356] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.096034s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/9/items I0420 19:26:21.286808 135416661391104 atomicity.py:608] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/9/items I0420 19:26:21.722249 135416661391104 atomicity.py:608] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/9 I0420 19:26:23.021513 135416661391104 atomicity.py:790] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/9`. I0420 19:26:23.022346 135416661391104 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 51.055790s. directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/9 I0420 19:26:23.023784 135416661391104 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 50.076249s. I0420 19:26:23.023955 135413448558336 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0420 19:26:23.024023 135413448558336 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0420 19:26:23.024100 135413448558336 checkpoint_manager.py:2133] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0420 19:26:23.025899 135413448558336 checkpoint_manager.py:2142] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0420 19:26:23.026096 135546001930048 checkpoint_manager.py:2036] [process=0][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9. I0420 19:26:23.026257 135546001930048 checkpoint_manager.py:2013] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0420 19:26:23.027059 135546001930048 metric_logger.py:185] completed step: 9, seconds: 0.230, TFLOP/s/device: 59.196, Tokens/s/device: 8922.698, total_weights: 65536, loss: 8.175 Per train step: Total TFLOPs: 13.59 split as 93.93% learnable weight flops and 6.07% attention flops XPK End: Mon Apr 20 19:26:35 UTC 2026 EXIT_CODE=0